skills/zjy365/skills/mole/Gen Agent Trust Hub

mole

Pass

Audited by Gen Agent Trust Hub on Aug 21, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes the mo executable for disk analysis and maintenance tasks.\n
  • Preflight checks verify the environment using uname -s and command -v mo to ensure it is running on a supported platform (macOS).\n
  • Commands are routed based on user intent (e.g., mo analyze, mo clean --dry-run) utilizing the tool's native capabilities.\n- [SAFE]: The skill implements robust safety protocols that align with security best practices.\n
  • Mandatory dry-run previews are required before any destructive action is taken.\n
  • An explicit 'Confirmation Contract' requires the user to approve a detailed plan in a separate turn before execution.\n
  • The skill includes a 'Revalidate Before Execution' protocol to confirm that candidate paths and environment states have not changed since the preview.\n
  • It explicitly prohibits the use of sudo and raw deletion commands (e.g., rm), relying entirely on the Mole tool's internal safety mechanisms.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 21, 2026, 06:30 AM
Security Audit — agent-trust-hub — mole